def t_dexp_dmu(self, model, Y, Y_metadata):
print("\n{}".format(inspect.stack()[0][3]))
#Make mu and var (marginal means and variances of q(f)) draws from a GP
k = GPy.kern.RBF(1).K(np.linspace(0, 1, Y.shape[0])[:, None])
L = GPy.util.linalg.jitchol(k)
mu = L.dot(np.random.randn(*Y.shape))
#Variance must be positive
var = np.abs(L.dot(np.random.randn(*Y.shape))) + 0.01
expectation = functools.partial(model.variational_expectations,
Y=Y,
v=var,
gh_points=None,
Y_metadata=Y_metadata)
#Function to get the nth returned value
def F(mu):
return expectation(m=mu)[0]
def dmu(mu):
return expectation(m=mu)[1]
grad = GradientChecker(F, dmu, mu.copy(), 'm')
grad.randomize()
print(grad)
print(model)
assert grad.checkgrad(verbose=1)
def t_d3transf_df3(self, transformation, f):
print "\n{}".format(inspect.stack()[0][3])
grad = GradientChecker(transformation.d2transf_df2,
transformation.d3transf_df3, f, 'f')
grad.randomize()
grad.checkgrad(verbose=1)
assert grad.checkgrad()
def dparam_checkgrad(func,
dfunc,
params,
params_names,
args,
constraints=None,
randomize=False,
verbose=False):
"""
checkgrad expects a f: R^N -> R^1 and df: R^N -> R^N
However if we are holding other parameters fixed and moving something else
We need to check the gradient of each of the fixed parameters
(f and y for example) seperately, whilst moving another parameter.
Otherwise f: gives back R^N and
df: gives back R^NxM where M is
The number of parameters and N is the number of data
Need to take a slice out from f and a slice out of df
"""
print "\n{} likelihood: {} vs {}".format(func.im_self.__class__.__name__,
func.__name__, dfunc.__name__)
partial_f = dparam_partial(func, *args)
partial_df = dparam_partial(dfunc, *args)
gradchecking = True
zipped_params = zip(params, params_names)
for param_ind, (param_val, param_name) in enumerate(zipped_params):
#Check one parameter at a time, make sure it is 2d (as some gradients only return arrays) then strip out the parameter
fnum = np.atleast_2d(partial_f(param_val,
param_name))[:, param_ind].shape[0]
dfnum = np.atleast_2d(partial_df(param_val,
param_name))[:, param_ind].shape[0]
for fixed_val in range(dfnum):
#dlik and dlik_dvar gives back 1 value for each
f_ind = min(fnum, fixed_val + 1) - 1
print "fnum: {} dfnum: {} f_ind: {} fixed_val: {}".format(
fnum, dfnum, f_ind, fixed_val)
#Make grad checker with this param moving, note that set_params is NOT being called
#The parameter is being set directly with __setattr__
#Check only the parameter and function value we wish to check at a time
grad = GradientChecker(
lambda p_val: np.atleast_2d(partial_f(p_val, param_name))[
f_ind, param_ind], lambda p_val: np.atleast_2d(
partial_df(p_val, param_name))[fixed_val, param_ind],
param_val, [param_name])
if constraints is not None:
for constrain_param, constraint in constraints:
if grad.grep_param_names(constrain_param):
constraint(constrain_param, grad)
else:
print "parameter didn't exist"
print constrain_param, " ", constraint
if randomize:
grad.randomize()
if verbose:
print grad
grad.checkgrad(verbose=1)
if not grad.checkgrad(verbose=True):
gradchecking = False
return gradchecking
def t_dexp_dmu(self, model, Y, Y_metadata):
print("\n{}".format(inspect.stack()[0][3]))
# Make mu and var (marginal means and variances of q(f)) draws from a GP
k = GPy.kern.RBF(1).K(np.linspace(0, 1, Y.shape[0])[:, None])
L = GPy.util.linalg.jitchol(k)
mu = L.dot(np.random.randn(*Y.shape))
# Variance must be positive
var = np.abs(L.dot(np.random.randn(*Y.shape))) + 0.01
expectation = functools.partial(
model.variational_expectations, Y=Y, v=var, gh_points=None, Y_metadata=Y_metadata
)
# Function to get the nth returned value
def F(mu):
return expectation(m=mu)[0]
def dmu(mu):
return expectation(m=mu)[1]
grad = GradientChecker(F, dmu, mu.copy(), "m")
grad.randomize()
print(grad)
print(model)
assert grad.checkgrad(verbose=1)
def t_dtransf_df(self, transformation, f):
print("\n{}".format(inspect.stack()[0][3]))
grad = GradientChecker(transformation.transf,
transformation.dtransf_df, f, 'f')
grad.randomize()
grad.checkgrad(verbose=1)
assert grad.checkgrad()
def t_d3logpdf_df3(self, model, Y, f, Y_metadata):
print("\n{}".format(inspect.stack()[0][3]))
d2logpdf_df2 = functools.partial(model.d2logpdf_df2, y=Y, Y_metadata=Y_metadata)
d3logpdf_df3 = functools.partial(model.d3logpdf_df3, y=Y, Y_metadata=Y_metadata)
grad = GradientChecker(d2logpdf_df2, d3logpdf_df3, f.copy(), "g")
grad.randomize()
print(model)
assert grad.checkgrad(verbose=1)
def t_d3logpdf_df3(self, model, Y, f):
print "\n{}".format(inspect.stack()[0][3])
d2logpdf_df2 = functools.partial(model.d2logpdf_df2, y=Y)
d3logpdf_df3 = functools.partial(model.d3logpdf_df3, y=Y)
grad = GradientChecker(d2logpdf_df2, d3logpdf_df3, f.copy(), 'g')
grad.randomize()
print model
assert grad.checkgrad(verbose=1)
def t_d3logpdf_df3(self, model, Y, f):
print "\n{}".format(inspect.stack()[0][3])
d2logpdf_df2 = functools.partial(model.d2logpdf_df2, y=Y)
d3logpdf_df3 = functools.partial(model.d3logpdf_df3, y=Y)
grad = GradientChecker(d2logpdf_df2, d3logpdf_df3, f.copy(), 'g')
grad.randomize()
print model
assert grad.checkgrad(verbose=1)
def t_dlogpdf_df(self, model, Y, f):
print "\n{}".format(inspect.stack()[0][3])
self.description = "\n{}".format(inspect.stack()[0][3])
logpdf = functools.partial(model.logpdf, y=Y)
dlogpdf_df = functools.partial(model.dlogpdf_df, y=Y)
grad = GradientChecker(logpdf, dlogpdf_df, f.copy(), 'g')
grad.randomize()
print model
assert grad.checkgrad(verbose=1)
def t_dlogpdf_df(self, model, Y, f):
print "\n{}".format(inspect.stack()[0][3])
self.description = "\n{}".format(inspect.stack()[0][3])
logpdf = functools.partial(model.logpdf, y=Y)
dlogpdf_df = functools.partial(model.dlogpdf_df, y=Y)
grad = GradientChecker(logpdf, dlogpdf_df, f.copy(), 'g')
grad.randomize()
print model
assert grad.checkgrad(verbose=1)
def t_dlogpdf_df(self, model, Y, f, Y_metadata):
print("\n{}".format(inspect.stack()[0][3]))
self.description = "\n{}".format(inspect.stack()[0][3])
logpdf = functools.partial(np.sum(model.logpdf), y=Y, Y_metadata=Y_metadata)
dlogpdf_df = functools.partial(model.dlogpdf_df, y=Y, Y_metadata=Y_metadata)
grad = GradientChecker(logpdf, dlogpdf_df, f.copy(), "g")
grad.randomize()
print(model)
assert grad.checkgrad(verbose=1)
def t_d2logpdf_df2(self, model, Y, f, Y_metadata):
print("\n{}".format(inspect.stack()[0][3]))
dlogpdf_df = functools.partial(model.dlogpdf_df,
y=Y,
Y_metadata=Y_metadata)
d2logpdf_df2 = functools.partial(model.d2logpdf_df2,
y=Y,
Y_metadata=Y_metadata)
grad = GradientChecker(dlogpdf_df, d2logpdf_df2, f.copy(), 'g')
grad.randomize()
print(model)
assert grad.checkgrad(verbose=1)
def t_dlogpdf_df(self, model, Y, f, Y_metadata):
print("\n{}".format(inspect.stack()[0][3]))
self.description = "\n{}".format(inspect.stack()[0][3])
logpdf = functools.partial(np.sum(model.logpdf),
y=Y,
Y_metadata=Y_metadata)
dlogpdf_df = functools.partial(model.dlogpdf_df,
y=Y,
Y_metadata=Y_metadata)
grad = GradientChecker(logpdf, dlogpdf_df, f.copy(), 'g')
grad.randomize()
print(model)
assert grad.checkgrad(verbose=1)
def t_dlogpdf_dlink(self, model, Y, f, link_f_constraints):
print "\n{}".format(inspect.stack()[0][3])
logpdf = functools.partial(model.logpdf_link, y=Y)
dlogpdf_dlink = functools.partial(model.dlogpdf_dlink, y=Y)
grad = GradientChecker(logpdf, dlogpdf_dlink, f.copy(), 'g')
#Apply constraints to link_f values
for constraint in link_f_constraints:
constraint('g', grad)
grad.randomize()
print grad
grad.checkgrad(verbose=1)
assert grad.checkgrad()
def t_d3logpdf_dlink3(self, model, Y, f, link_f_constraints):
print "\n{}".format(inspect.stack()[0][3])
d2logpdf_dlink2 = functools.partial(model.d2logpdf_dlink2, y=Y)
d3logpdf_dlink3 = functools.partial(model.d3logpdf_dlink3, y=Y)
grad = GradientChecker(d2logpdf_dlink2, d3logpdf_dlink3, f.copy(), 'g')
#Apply constraints to link_f values
for constraint in link_f_constraints:
constraint('g', grad)
grad.randomize()
grad.checkgrad(verbose=1)
print grad
assert grad.checkgrad()
def t_d3logpdf_dlink3(self, model, Y, f, Y_metadata, link_f_constraints):
print("\n{}".format(inspect.stack()[0][3]))
d2logpdf_dlink2 = functools.partial(model.d2logpdf_dlink2, y=Y, Y_metadata=Y_metadata)
d3logpdf_dlink3 = functools.partial(model.d3logpdf_dlink3, y=Y, Y_metadata=Y_metadata)
grad = GradientChecker(d2logpdf_dlink2, d3logpdf_dlink3, f.copy(), "g")
# Apply constraints to link_f values
for constraint in link_f_constraints:
constraint("g", grad)
grad.randomize()
print(grad)
print(model)
assert grad.checkgrad(verbose=1)
def dparam_checkgrad(func, dfunc, params, params_names, args, constraints=None, randomize=False, verbose=False):
"""
checkgrad expects a f: R^N -> R^1 and df: R^N -> R^N
However if we are holding other parameters fixed and moving something else
We need to check the gradient of each of the fixed parameters
(f and y for example) seperately, whilst moving another parameter.
Otherwise f: gives back R^N and
df: gives back R^NxM where M is
The number of parameters and N is the number of data
Need to take a slice out from f and a slice out of df
"""
print "\n{} likelihood: {} vs {}".format(func.im_self.__class__.__name__,
func.__name__, dfunc.__name__)
partial_f = dparam_partial(func, *args)
partial_df = dparam_partial(dfunc, *args)
gradchecking = True
zipped_params = zip(params, params_names)
for param_ind, (param_val, param_name) in enumerate(zipped_params):
#Check one parameter at a time, make sure it is 2d (as some gradients only return arrays) then strip out the parameter
fnum = np.atleast_2d(partial_f(param_val, param_name))[:, param_ind].shape[0]
dfnum = np.atleast_2d(partial_df(param_val, param_name))[:, param_ind].shape[0]
for fixed_val in range(dfnum):
#dlik and dlik_dvar gives back 1 value for each
f_ind = min(fnum, fixed_val+1) - 1
print "fnum: {} dfnum: {} f_ind: {} fixed_val: {}".format(fnum, dfnum, f_ind, fixed_val)
#Make grad checker with this param moving, note that set_params is NOT being called
#The parameter is being set directly with __setattr__
#Check only the parameter and function value we wish to check at a time
grad = GradientChecker(lambda p_val: np.atleast_2d(partial_f(p_val, param_name))[f_ind, param_ind],
lambda p_val: np.atleast_2d(partial_df(p_val, param_name))[fixed_val, param_ind],
param_val, [param_name])
if constraints is not None:
for constrain_param, constraint in constraints:
if grad.grep_param_names(constrain_param):
constraint(constrain_param, grad)
else:
print "parameter didn't exist"
print constrain_param, " ", constraint
if randomize:
grad.randomize()
if verbose:
print grad
grad.checkgrad(verbose=1)
if not grad.checkgrad(verbose=True):
gradchecking = False
return gradchecking
def test_gaussian_d2logpdf_df2_2(self):
print("\n{}".format(inspect.stack()[0][3]))
self.Y = None
self.N = 2
self.D = 1
self.X = np.linspace(0, self.D, self.N)[:, None]
self.real_std = 0.2
noise = np.random.randn(*self.X.shape) * self.real_std
self.Y = np.sin(self.X * 2 * np.pi) + noise
self.f = np.random.rand(self.N, 1)
dlogpdf_df = functools.partial(self.gauss.dlogpdf_df, y=self.Y)
d2logpdf_df2 = functools.partial(self.gauss.d2logpdf_df2, y=self.Y)
grad = GradientChecker(dlogpdf_df, d2logpdf_df2, self.f.copy(), 'g')
grad.randomize()
self.assertTrue(grad.checkgrad(verbose=1))
def test_gaussian_d2logpdf_df2_2(self):
print("\n{}".format(inspect.stack()[0][3]))
self.Y = None
self.N = 2
self.D = 1
self.X = np.linspace(0, self.D, self.N)[:, None]
self.real_std = 0.2
noise = np.random.randn(*self.X.shape) * self.real_std
self.Y = np.sin(self.X * 2 * np.pi) + noise
self.f = np.random.rand(self.N, 1)
dlogpdf_df = functools.partial(self.gauss.dlogpdf_df, y=self.Y)
d2logpdf_df2 = functools.partial(self.gauss.d2logpdf_df2, y=self.Y)
grad = GradientChecker(dlogpdf_df, d2logpdf_df2, self.f.copy(), "g")
grad.randomize()
self.assertTrue(grad.checkgrad(verbose=1))
def t_d2logpdf_dlink2(self, model, Y, f, Y_metadata, link_f_constraints):
print("\n{}".format(inspect.stack()[0][3]))
dlogpdf_dlink = functools.partial(model.dlogpdf_dlink,
y=Y,
Y_metadata=Y_metadata)
d2logpdf_dlink2 = functools.partial(model.d2logpdf_dlink2,
y=Y,
Y_metadata=Y_metadata)
grad = GradientChecker(dlogpdf_dlink, d2logpdf_dlink2, f.copy(), 'g')
#Apply constraints to link_f values
for constraint in link_f_constraints:
constraint('g', grad)
grad.randomize()
print(grad)
print(model)
assert grad.checkgrad(verbose=1)
def dparam_checkgrad(func, dfunc, params, args, constraints=None, randomize=False, verbose=False):
"""
checkgrad expects a f: R^N -> R^1 and df: R^N -> R^N
However if we are holding other parameters fixed and moving something else
We need to check the gradient of each of the fixed parameters
(f and y for example) seperately, whilst moving another parameter.
Otherwise f: gives back R^N and
df: gives back R^NxM where M is
The number of parameters and N is the number of data
Need to take a slice out from f and a slice out of df
"""
#print "\n{} likelihood: {} vs {}".format(func.im_self.__class__.__name__,
#func.__name__, dfunc.__name__)
partial_f = dparam_partial(func, *args)
partial_df = dparam_partial(dfunc, *args)
gradchecking = True
for param in params:
fnum = np.atleast_1d(partial_f(param)).shape[0]
dfnum = np.atleast_1d(partial_df(param)).shape[0]
for fixed_val in range(dfnum):
#dlik and dlik_dvar gives back 1 value for each
f_ind = min(fnum, fixed_val+1) - 1
print "fnum: {} dfnum: {} f_ind: {} fixed_val: {}".format(fnum, dfnum, f_ind, fixed_val)
#Make grad checker with this param moving, note that set_params is NOT being called
#The parameter is being set directly with __setattr__
grad = GradientChecker(lambda x: np.atleast_1d(partial_f(x))[f_ind],
lambda x : np.atleast_1d(partial_df(x))[fixed_val],
param, 'p')
#This is not general for more than one param...
if constraints is not None:
for constraint in constraints:
constraint('p', grad)
if randomize:
grad.randomize()
if verbose:
print grad
grad.checkgrad(verbose=1)
if not grad.checkgrad():
gradchecking = False
return gradchecking
def dparam_checkgrad(func, dfunc, params, params_names, args, constraints=None, randomize=False, verbose=False):
"""
checkgrad expects a f: R^N -> R^1 and df: R^N -> R^N
However if we are holding other parameters fixed and moving something else
We need to check the gradient of each of the fixed parameters
(f and y for example) seperately, whilst moving another parameter.
Otherwise f: gives back R^N and
df: gives back R^NxM where M is
The number of parameters and N is the number of data
Need to take a slice out from f and a slice out of df
"""
print("\n{} likelihood: {} vs {}".format(func.__self__.__class__.__name__, func.__name__, dfunc.__name__))
partial_f = dparam_partial(func, *args)
partial_df = dparam_partial(dfunc, *args)
gradchecking = True
zipped_params = zip(params, params_names)
for param_ind, (param_val, param_name) in enumerate(zipped_params):
# Check one parameter at a time, make sure it is 2d (as some gradients only return arrays) then strip out the parameter
f_ = partial_f(param_val, param_name)
df_ = partial_df(param_val, param_name)
# Reshape it such that we have a 3d matrix incase, that is we want it (?, N, D) regardless of whether ? is num_params or not
f_ = f_.reshape(-1, f_.shape[0], f_.shape[1])
df_ = df_.reshape(-1, f_.shape[0], f_.shape[1])
# Get the number of f and number of dimensions
fnum = f_.shape[-2]
fdim = f_.shape[-1]
dfnum = df_.shape[-2]
for fixed_val in range(dfnum):
# dlik and dlik_dvar gives back 1 value for each
f_ind = min(fnum, fixed_val + 1) - 1
print("fnum: {} dfnum: {} f_ind: {} fixed_val: {}".format(fnum, dfnum, f_ind, fixed_val))
# Make grad checker with this param moving, note that set_params is NOT being called
# The parameter is being set directly with __setattr__
# Check only the parameter and function value we wish to check at a time
# func = lambda p_val, fnum, fdim, param_ind, f_ind, param_ind: partial_f(p_val, param_name).reshape(-1, fnum, fdim)[param_ind, f_ind, :]
# dfunc_dparam = lambda d_val, fnum, fdim, param_ind, fixed_val: partial_df(d_val, param_name).reshape(-1, fnum, fdim)[param_ind, fixed_val, :]
# First we reshape the output such that it is (num_params, N, D) then we pull out the relavent parameter-findex and checkgrad just this index at a time
func = lambda p_val: partial_f(p_val, param_name).reshape(-1, fnum, fdim)[param_ind, f_ind, :]
dfunc_dparam = lambda d_val: partial_df(d_val, param_name).reshape(-1, fnum, fdim)[param_ind, fixed_val, :]
grad = GradientChecker(func, dfunc_dparam, param_val, [param_name])
if constraints is not None:
for constrain_param, constraint in constraints:
if grad.grep_param_names(constrain_param):
constraint(constrain_param, grad)
else:
print("parameter didn't exist")
print(constrain_param, " ", constraint)
if randomize:
grad.randomize()
if verbose:
print(grad)
grad.checkgrad(verbose=1)
if not grad.checkgrad(verbose=True):
gradchecking = False
if not grad.checkgrad(verbose=True):
gradchecking = False
return gradchecking
def t_d3transf_df3(self, transformation, f):
print "\n{}".format(inspect.stack()[0][3])
grad = GradientChecker(transformation.d2transf_df2, transformation.d3transf_df3, f, 'f')
grad.randomize()
grad.checkgrad(verbose=1)
assert grad.checkgrad()
def dparam_checkgrad(func,
dfunc,
params,
params_names,
args,
constraints=None,
randomize=False,
verbose=False):
"""
checkgrad expects a f: R^N -> R^1 and df: R^N -> R^N
However if we are holding other parameters fixed and moving something else
We need to check the gradient of each of the fixed parameters
(f and y for example) seperately, whilst moving another parameter.
Otherwise f: gives back R^N and
df: gives back R^NxM where M is
The number of parameters and N is the number of data
Need to take a slice out from f and a slice out of df
"""
print("\n{} likelihood: {} vs {}".format(func.__self__.__class__.__name__,
func.__name__, dfunc.__name__))
partial_f = dparam_partial(func, *args)
partial_df = dparam_partial(dfunc, *args)
gradchecking = True
zipped_params = zip(params, params_names)
for param_ind, (param_val, param_name) in enumerate(zipped_params):
#Check one parameter at a time, make sure it is 2d (as some gradients only return arrays) then strip out the parameter
f_ = partial_f(param_val, param_name)
df_ = partial_df(param_val, param_name)
#Reshape it such that we have a 3d matrix incase, that is we want it (?, N, D) regardless of whether ? is num_params or not
f_ = f_.reshape(-1, f_.shape[0], f_.shape[1])
df_ = df_.reshape(-1, f_.shape[0], f_.shape[1])
#Get the number of f and number of dimensions
fnum = f_.shape[-2]
fdim = f_.shape[-1]
dfnum = df_.shape[-2]
for fixed_val in range(dfnum):
#dlik and dlik_dvar gives back 1 value for each
f_ind = min(fnum, fixed_val + 1) - 1
print("fnum: {} dfnum: {} f_ind: {} fixed_val: {}".format(
fnum, dfnum, f_ind, fixed_val))
#Make grad checker with this param moving, note that set_params is NOT being called
#The parameter is being set directly with __setattr__
#Check only the parameter and function value we wish to check at a time
#func = lambda p_val, fnum, fdim, param_ind, f_ind, param_ind: partial_f(p_val, param_name).reshape(-1, fnum, fdim)[param_ind, f_ind, :]
#dfunc_dparam = lambda d_val, fnum, fdim, param_ind, fixed_val: partial_df(d_val, param_name).reshape(-1, fnum, fdim)[param_ind, fixed_val, :]
#First we reshape the output such that it is (num_params, N, D) then we pull out the relavent parameter-findex and checkgrad just this index at a time
func = lambda p_val: partial_f(p_val, param_name).reshape(
-1, fnum, fdim)[param_ind, f_ind, :]
dfunc_dparam = lambda d_val: partial_df(d_val, param_name).reshape(
-1, fnum, fdim)[param_ind, fixed_val, :]
grad = GradientChecker(func, dfunc_dparam, param_val, [param_name])
if constraints is not None:
for constrain_param, constraint in constraints:
if grad.grep_param_names(constrain_param):
constraint(constrain_param, grad)
else:
print("parameter didn't exist")
print(constrain_param, " ", constraint)
if randomize:
grad.randomize()
if verbose:
print(grad)
grad.checkgrad(verbose=1)
if not grad.checkgrad(verbose=True):
gradchecking = False
if not grad.checkgrad(verbose=True):
gradchecking = False
return gradchecking
